Research map / Graph learning / GNN architectures
Topology-Aware Graph Pooling Networks
Hongyang Gao, Yi Liu, Shuiwang Ji
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add TAP layers (ablation) | TAPNet · PTC | accuracy (%): 70.6 -> 73.4 | Improved |
| add TAP layers (ablation) | TAPNet · IMDB-MULTI | accuracy (%): 52.1 -> 56.2 | Improved |
| add TAP layers (ablation) | TAPNet · REDDIT-BINARY | accuracy (%): 91.0 -> 94.6 | Improved |
| add local voting term (ablation) | TAP layer · PTC | accuracy (%): 72.1 -> 73.4 | Improved |
| add local voting term (ablation) | TAP layer · REDDIT-BINARY | accuracy (%): 93.2 -> 94.6 | Improved |
| add global voting term (ablation) | TAP layer · PTC | accuracy (%): 72.7 -> 73.4 | Improved |
| add global voting term (ablation) | TAP layer · REDDIT-BINARY | accuracy (%): 94.1 -> 94.6 | Improved |
| add graph connection term (ablation) | TAP layer · PTC | accuracy (%): 73.1 -> 73.4 | Improved |
| add graph connection term (ablation) | TAP layer · REDDIT-BINARY | accuracy (%): 94.2 -> 94.6 | Improved |
| add auxiliary link prediction objective (ablation) | TAPNet · REDDIT-BINARY | accuracy (%): 94.6 -> 94.8 | Improved |
| add local and global voting to TAP scoring (ablation) | TAPNet · REDDIT-BINARY | accuracy (%): 91.5 -> 94.6 | Improved |
| use TAP layer instead of top-k pooling | TAPNet architecture · PTC | accuracy (%): 71.5 -> 73.4 | Improved |
Closest papers by meaning
- Geometry-Aware Edge Pooling for Graph Neural Networks (NeurIPS 2025 · similarity 0.90)
- SkipPool: Improved Sparse Hierarchical Graph Pooling with Differentiable Exploration (AAAI 2025 · similarity 0.89)
- Boosting Graph Pooling with Persistent Homology (NeurIPS 2024 · similarity 0.89)
- MaxCutPool: differentiable feature-aware Maxcut for pooling in graph neural networks (ICLR 2025 · similarity 0.87)
- Graph Parsing Networks (ICLR 2024 · similarity 0.87)
- Fair Graph U-Net: A Fair Graph Learning Framework Integrating Group and Individual Awareness (AAAI 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
